稳健统计方法及t检验法在活菌计数能力比对中的应用

Application of Robust Statistic Method and T Test of Proficiency Testing for Live Bacteria Counting

关键词:  活菌计数  能力比对  稳健统计技术  t检验法

Key words:  Live Bacteria Counting  Proficiency Test  Robust Statistical Techniques  T Test

摘要:

为了解检验员活菌计数能力，寻求更简便、准确的计数方法，8名检验员同时同环境对同一细菌培养物采用常量法和微量法进行活菌计数，采用稳健统计技术对不同人员相同方法的计数结果进行统计，将检验结果分为“满意”、“可疑”和“不满意”；采用t检验法对使用不同方法对同一样品的计数结果进行统计，将结果分为“差异显著”和“差异不显著”。用常量法进行检测，有6名检验员的检验结果全部满意，而微量法的结果中只有2名检验员获得了全部满意的结果。两种方法的检验结果差异显著。比对结果表明，用微量法进行检验，人员之间的差异更大。

Abstract:

The research is aim to comprehensively grasp inspectors' ability of live bacteria counting, seek a more simple and accurate counting method. Each inspector counted bacteria in the same culture using two methods of constant and micro methods in the same environment at the same time. Then robust statistical techniques to the analysis of Z score values were used to statistically analysized live bacteria counting results of different inspectors using same method. Different inspectors' test results were evaluated as "satisfactory", "suspicious" or "unsatisfactory". T test was used to statistically analysized live bacteria counting results of different methods but same sample. Test results using different methods were evaluated as " significant difference " and "no significant difference". Three kinds of bacteria were detected by constant method, six inspectors’ test results are all satisfactory. But, only two inspectors’ test results are all satisfactory using micro method. Besides, the differences between results using two methods were significant. The test result differences among inspectors using micro method were more significant than that using constant method.